Electric power transmission has traditionally utilized overhead power lines suspended from pylons. To transport large amounts of power over long distances with low loss, it is necessary to use high-voltage direct current (HVDC) and low resistance (thus thick) conductors. Using overhead conductors in this application requires tall, sturdy towers. Particularly in urban or scenic areas, there is substantial resistance to allowing such towers to be built. Alternatively, power lines can be insulated and placed underground. The three basic prior art technologies to do this are (non-superconducting) underground cables, gas insulated lines (GIL), and superconducting underground cables.
An individual underground cable is limited in its diameter to that which can be coiled on a spool and transported to the point of installation. This diameter limit in turn limits the thickness of the electrical insulation and thickness of the conductor in such a wire. The conductor size (cross-sectional area for DC) determines its resistance, which sets the amount of heat generated (I2R) by the current it carries. The insulation must be thick enough to prevent electrical breakdown at the intended working voltage. The higher the voltage, the more power that can be transported at a given current (and thus a given level of heating), but the thicker the required insulation. A thicker insulation causes a higher temperature for a given level of waste heat generation. The insulation will fail if its upper temperature limit is exceeded, so these factors conspire to set an upper limit of around 1.1 GW of power that can be transported per state-of-the-art underground cable. Distances must be kept relatively short, as losses are relatively high (19% for 1,000 km is typical today). Unlike overhead lines, where capacity scales approximately with the square of voltage, the need to move the waste heat out of a cable through the insulation means that capacity of a cable scales more nearly linearly with the voltage.
GIL, on the other hand, uses rigid conductor sections, and thus is not limited by the requirement of needing sufficient flexibility for the conductor to be coiled onto a transportable spool. Pressurized sulfur hexafluoride (SF6) gas mixed with pressurized nitrogen gas (N2) provide the insulation. Because of gas convection, GIL is far better at moving waste heat out of the conductor to the outer perimeter of the insulation, which in GIL is usually a metal pipe. Although this technology can scale up to larger capacities and keep losses small by using large conductor sizes, the requirement to manage a pressurized gas along with the fact that SF6 is a very potent greenhouse gas (22,800 times as bad as CO2 over 100 years, and thus suitable precautions must be taken to prevent accidental discharges), make this technology comparatively expensive.
Even “high temperature” superconducting (HTS) power transmission lines require cryogenic cooling (liquid nitrogen temperature) to maintain the conductors in a superconducting (zero resistance) state. The refrigeration equipment adds a power loss penalty and creates a reliability issue for such power lines, since they must go out of service if the refrigeration falters.
It is a widely held belief within the electric power industry that long distance transmission lines cannot be made out of short (15 meter, easily truck-transportable sections) because the cumulative reliability for ˜130,000 splices (λ130,000, where λ is the reliability of one splice over a given time) needed for a 1,000 km transmission line would be unacceptably low. (For example with a single splice reliability of 99.999% over a year, the above 1,000 km transmission line as a whole would have a reliability of only 26% over the same year—or a 74% probability of failing). Thus an unacceptable aggregate reliability is a problem with the splicing technology currently deployed for underground cables.
To illustrate, splicing of underground power cables using current methods as shown by Tatsuya Nagata, et al., in “Flexible Joint for 275 kV XLPE Cable”, and Akira Suzuki, et al., in “Installation of the World's First 500-kV XLPE Cable with Intermediate Joints” first involves stripping and exposing the multiple layers of the cable, then butt welding or soldering the numerous individual conductor strands that make up the complete conductor. Wrapping the conductor with semiconductive tape, and then casting or extruding insulating polymers over the electrical connection follows this.
Gold's U.S. Pat. No. 4,270,021 shows one such method. Any imperfections in this process [such as Contaminants, Protrusions, or Voids (referred to as CPV's in the industry) in excess of a few microns, whether the contaminants be conductive or insulative] can result in premature failure by electrical breakdown, per Advances in High Voltage Engineering, page 494. Thus a clean room is assembled around each splice out in the field, and equipment is used to detect the presence of microscopic impurities in the insulation resin, followed by sensitive inspection equipment (such as X-ray) to gain assurance that the completed splice will be reliable. The next step is to cover the insulation with semiconductive tape, followed by cushioning, sheathing, and jacketing layers. The whole existing process is labor intensive, time consuming, and complex, and constitutes a weak point in transmission system reliability even with only a small number of splices.
Many methods have existed to form a splice between aluminum and copper conductors. These methods are relatively complex and thus costly and a source of unreliability.
When splicing rigid conductors in GIL transmission lines, one must allow for the expansion and contraction of the conductor due to changes in its temperature. Over typical temperature excursions, a rigid aluminum conductor that is nominally 15 meters in length may contract or expand by nearly 2.5 cm. This can be caused by both changes in the environmental temperature and the current (thus I2R loss) that the line is carrying. With wires or flexible cables this expansion or contraction can be taken up by small changes in droop or snaking of the wire or cable, but not so with rigid conductors.
One GIL electrical expansion joint method that is in current use involves a sliding mechanical contact between overlapping sections of adjoining conductor sections. This method uses spring force to assure that the contact pressure is sufficient to achieve low resistance, and thus loss. Since this process inherently involves sliding metal-to-metal contact under pressure, wear will occur. This wear will produce a conductive metallic dust. Since a gas insulates GIL transmission lines, the dust may freely move about within this medium. The insulating gas contains an electric field. These particulates modify the electric field around themselves, leading to regions of increased electrical stress, which can initiate electrical breakdown. Thus GIL designs include particulate traps to hopefully gather these particles and remove them from areas where they could induce a failure.
When particulates become electrically charged they will be attracted to and move toward an oppositely charged surface. In an HVDC GIL environment, the particles generated from sliding contactors and other sources can and do go back and forth between the conductor and the outer conduit as their charge changes each time they touch a metal wall or conductor, and this allows the particles to bounce along and accumulate, which is particularly a problem only when GIL is deployed in an HVDC transmission mode. This bouncing dust problem is much less severe for GIL lines carrying AC power.
Transporting large amounts of power, even at relatively high efficiency, results in the generation of a significant amount of waste heat. For instance, to move 10 GW of power with a loss of 3% per 1,000 km (corresponding to the loss of the best conventional conductor-based power transmission lines built to date) creates 150 watts of waste heat per meter, per conductor. This is well above the maximum waste heat that can reliably be shed by underground conductors through soil, even if a cable that could carry 10 GW were available. Thus, a breakthrough on higher capacity underground passively cooled power lines will require higher efficiency than the best available power lines of today. Larger capacities and larger loss levels result in proportionally larger amounts of waste heat. Different dirt and soil types have varying (with moisture level, for instance) and relatively low thermal conductivities (0.06 to 3 W/m·° K), and coupling of waste heat that has been brought to the surface to be removed by the surrounding air can be hindered by vegetation and debris.
To summarize, the key technical obstacles for achieving commercially viable underground high-power electric transmission lines using conventional conductors are:                1. High voltage electrical insulation for the conductors        2. Removal of waste heat from the conductors        3. Accommodating thermal expansion and contraction of the conductors, insulation, splices, and housings        4. Making low loss, high current, highly reliable electrical splices in the field at low cost        5. Reliably insulating high voltage field-made splices at low cost        
Since misguided digging accounts for about one-half of the field failures (per Advances in High Voltage Engineering, page 491) seen with underground power cables, it is likewise desirable to make it extremely unlikely that an underground transmission line would be severed accidentally by digging or other construction activity. One prior art way to accomplish this is to install cables inside pipes (conduit); these are known as “pipe-type cables,” and actually pre-date directly buried high voltage cables. Pipe type cables make it far less likely for digging to damage a buried electric line. In present practice, pipe-type cables are installed between maintenance and repair vaults that are no more than about a kilometer apart (the maximum distance that can be pulled without a splice in a large cable).
Perhaps the most common examples of a high voltage, high current conductors that are not wires are the buss bars found in power plants and transformer yards. These are made of extruded aluminum and are air-insulated. These are not used for transmission outside of power plants, electric utility transformer and switching sites, and within the facilities of large industrial power users.
All power lines have some limitation on the minimum radius of curvature for each particular type of power line. In fact, there can be different limitations for minimum radius of curvature for shipping a conductor (where it is generally tightly wrapped on a reel for transport), compared to installation. For example, large pipe-type cables may come on a 3 meter diameter cable reel, but they cannot be pulled through a conduit with less than about a 30 meter radius of curvature. To achieve tighter turns than this, it is necessary to have a maintenance vault and a splice where the pipe-type cable turns a corner.